tatami_chunked
Helpers to create custom chunked tatami matrices
|
Options for data extraction from a CustomSparseChunkedMatrix
.
More...
#include <CustomSparseChunkedMatrix.hpp>
Public Attributes | |
std::size_t | maximum_cache_size = sanisizer::cap<std::size_t>(100000000) |
bool | require_minimum_cache = true |
bool | cache_subset = false |
Options for data extraction from a CustomSparseChunkedMatrix
.
std::size_t tatami_chunked::CustomSparseChunkedMatrixOptions::maximum_cache_size = sanisizer::cap<std::size_t>(100000000) |
Size of the in-memory cache in bytes. Larger caches improve access speed at the cost of memory usage. Small values may be ignored if require_minimum_cache
is true
.
bool tatami_chunked::CustomSparseChunkedMatrixOptions::require_minimum_cache = true |
Whether to automatically enforce a minimum size for the cache, regardless of maximum_cache_size
. This minimum is chosen to ensure that a single slab can be retained in memory, so that the same chunks are not repeatedly re-read when iterating over consecutive rows/columns of the matrix.
bool tatami_chunked::CustomSparseChunkedMatrixOptions::cache_subset = false |
Whether to only extract and cache a subset of elements along the target dimension when an oracle is available. This involves more overhead to determine which elements are needed but may improve access speed for chunking strategies that support partial extraction.